Curved Point Geometry 00 gauge

[IRSanderson]

Can anyone help with the problem in the attached image? I've tried to create the recommended combination of points and curves for Hornby and Peco Settrack, but I get the same (highlighted) error in Anyrail. The outer curve has R607 and R610 (ST 226 and 203) but that doesn't join up successfully. Do I need a different combination or is just a bug in the Anyrail software?

[Golden Eagle]

There's no bug in Anyrail - I get the same with SCARM. The problem is that most of the lower half of the outer curve is built from 2nd radius curves, while the upper half is built from 3rd radius and, while the inner fork of the turnout is 2nd radius,  the outer fork is tighter than 3rd radius (444mm rather than 505mm radius).

 

There may be enough "wiggle room" in the track and couplers to bring the track ends together. If not, my "fix" would be to replace the R610/R607 or ST203/ST226 between the point and 3rd radius curve with a suitable length of flex-track (R621 or SL100).

 

 

[ejstubbs]

I think you need a different combination of track pieces.  I found these images on a retailer's web site a few years back and kept them handy as a reference:

 

 

That said, I've tried putting both the Hornby and Peco curved point crossovers together in AnyRail and there is still a gap of ~5mm in each case:

 

 

(Note that I've used a standard crossover on straight track at each end of the curve to provide the baselines.)

 

I think both 'solutions' as recommended by the manufacturers depend on a bit of "udging" to make all the track sections connect.  (My recollection from days past when I tried transferring plans from the Peco Setrack track plan book in to Anyrail is you end up with one or more small gaps with virtually every one of the plans.)  Absorb the 'error' between a number of connections and it's unlikely to be particularly noticeable to the eye, though it might not help with reliable running.

 

Using flexi to fix the problem could end up being problematic in itself as flexi doesn't behave well when being curved down to Setrack radii.

Edited January 2, 2021 by ejstubbs

[Dragonfly]

The geometry quoted by Hornby, Peco, and various books isn't actually correct, weirdly.

 

The outer curve of a curved point is actually equivalent to a 67mm straight folloed by a 2nd radius curve.

 

67mm is also the distance separation between tracks in SetTrack.

 

This means you can get a standard track separation over a 90-degree angle, following each branch of the point with 2nd radius curves:

 

However, this does mean a 67mm straight would need to be cut to make a crossover, so not much help there! Probably explains why the manufacturers use such a bodge with their stated geometry.

Edited January 2, 2021 by Dragonfly

[KeithMacdonald]

Just playing in AnyRail. It might be my imagination but is there a smaller gap with this combination? Or enough "wiggle-room" in using a pair of R606's to close the gap?

 

Edited January 2, 2021 by KeithMacdonald
Typo fix

[Butler Henderson]

Whilst following the setrack practice to make joints in parallel lines adjacent  the above can be reduced in cost  by a couple of quid by replacing the two R643s / ST-227s that are joined together with a R606  / ST-225.

 

Late edit: noticed that track arrangement occurs twice unnecessarily while there are also two R608 / ST230s where a R609 / ST231 would fit.

Edited January 3, 2021 by Butler Henderson

[Dungrange]

3 hours ago, KeithMacdonald said:

Just playing in AnyRail. It might be my imagination but is there a smaller gap with this combination? Or enough "wiggle-room" in using a pair of R606's to close the gap?

 

 

Could one of the R606 (second radius curves) be replaced by an R608 (third radius curve)?  That would push it out a bit.  Alternatively, could it be replaced by a R643 (half second radius) and whatever the equivalent half third radius curve is?

[IRSanderson]

Thanks for all the helpful suggestions.

[KeithMacdonald]

16 minutes ago, Dungrange said:

Could one of the R606 (second radius curves) be replaced by an R608 (third radius curve)?  That would push it out a bit.  Alternatively, could it be replaced by a R643 (half second radius) and whatever the equivalent half third radius curve is?

 

You are quite right, using a R608 (third radius curve) does improve the geometry of the curve, but sadly it still seems to leave a gap, which looks like half of an R610?

 

 

Regarding a R643 (half second radius, 11.25 degrees), does Hornby make a half third radius?

 

[Dungrange]

10 minutes ago, KeithMacdonald said:

Regarding a R643 (half second radius, 11.25 degrees), does Hornby make a half third radius?

 

It looks like the answer is no, they don't.  

 

https://www.Hornby.com/media/pdf/Track-Geometry-PDF.pdf is the latest version of the image that @ejstubbs posted above.  While it adds fourth radius, it appears that second radius is still the only instance of an 11.25 degree curve.

 

It also looks like there must be a bit of 'artistic licence' in the Hornby drawings.  At the bottom end of the curved crossover (ie the bottom of the plan), the first, second and third radius curves look like they are at standard track centres as stated, but they don't actually look like they are parallel.

[Butler Henderson]

The bodge would be to use a R609 in place of the R606 and R608, partly notching the sleepers  on the underside do its radius can be tightened slightly and cut to fit. A solution using setrack parts without bodging then would be to insert a straight at the center of the curve with 4 R610s on the inner curve and. R600 on the outer curve.

Edited January 2, 2021 by Butler Henderson

[RobinofLoxley]

On my old layout, whch I have just taken up, I had two junctions using curved points as shown in the lower half of Keith Macdonald's post just above, with R608 in place of 2 x R643, all R609's on the outer loop otherwise. The only difference the r643's make is that they give you more placing possibilities of the points within the quarter circle. However, such a small distance disappears in practice due to the number of track joins involved, which are never 100% right as to angle and overall positioning. About 1mm space on the outer half track gaps fixes the discrepancy in theory, in practice both my junctions were 100% parallel in and out. Even if there's a real problem as shown, closing all the outer line tracks together doesn't cause a significant loss of gauge.

 

Would be interested to know if replacing the antiquated R612/3 with 8072/3 changed anything.

 

Pity I dont have my original track plan as I followed the original Hornby plan as far as I can recall.

Edited January 2, 2021 by RobinofLoxley

[DCB]

On 02/01/2021 at 11:54, IRSanderson said:

Can anyone help with the problem in the attached image? I've tried to create the recommended combination of points and curves for Hornby and Peco Settrack, but I get the same (highlighted) error in Anyrail. The outer curve has R607 and R610 (ST 226 and 203) but that doesn't join up successfully. Do I need a different combination or is just a bug in the Anyrail software?

Stop stressing.  The misalignment is too small to make any difference in the real world. 

The rail joints allow a bit of flexibility.   I cut webs between sleepers an tweak set track from 18 to 20" radius, chop bits off to make non standard lengths, reduce track centres from 60mm to 42mm but your problem should be well within tolerances for unmodified set track.   I don't like the mosaic of short rails, bottom right,  though... 

Edited January 3, 2021 by DavidCBroad

[RobinofLoxley]

In a related activity, I have just test laid the track for a junction leading to a station, so a complex of points. Anyrail showed a negligible single gap but actually I will have to cut two sections of rail each 13mm long in order to bridge two small gaps. Fortunately 13mm is just the length of a fishplate, so I am guaranteed that I can produce something stable, even if I finish up soldering the extra pieces in place.

[DCB]

On 03/01/2021 at 18:47, RobinofLoxley said:

In a related activity, I have just test laid the track for a junction leading to a station, so a complex of points. Anyrail showed a negligible single gap but actually I will have to cut two sections of rail each 13mm long in order to bridge two small gaps. Fortunately 13mm is just the length of a fishplate, so I am guaranteed that I can produce something stable, even if I finish up soldering the extra pieces in place.

Its a common issue, Anyrail uses a bespoke template for each type of point and the manufacturers aren't so good at keeping their products to the same dimensions over long periods .  In my recent experience  the Peco 3ft radius point from 2015 does not fit the space previously occupied by a circa 1980 3ft radius point. later one is a couple of mm bigger, well maybe 6 overall. enough to be a darned nuisance.   What is a real nuisance id rails under about 50mm long, almost impossible to lay straight and level and my track testing device, a Hornby T9 always derails when they get a mm out of line vertically or horizontally.  as It has almost no flanges on its traction tyred wheels.

[KeithMacdonald]

On 02/01/2021 at 18:05, Dungrange said:

 

It looks like the answer is no, they don't.  

 

https://www.Hornby.com/media/pdf/Track-Geometry-PDF.pdf is the latest version of the image that @ejstubbs posted above.  While it adds fourth radius, it appears that second radius is still the only instance of an 11.25 degree curve.

 

It also looks like there must be a bit of 'artistic licence' in the Hornby drawings.  At the bottom end of the curved crossover (ie the bottom of the plan), the first, second and third radius curves look like they are at standard track centres as stated, but they don't actually look like they are parallel.

 

Thanks for that updated PDF.

 

There is one other 11.25 degree curve, R628, curve radius 85.2cm, way beyond the 4th Radius, but not relevant to this scenario at all?

 

[RobinofLoxley]

R628 is shown as simply being the companion track for a setrack Y-point, returning the track to parallel and in gauge. R643 returns the tracks to parallel but rather closer together.

 

I found R628's very useful in general track laying, and finished up with a fair number on my previous layout as they complement curved points when  used in other scenarios than just creating a cross-over in curved track as shown above.